Deep reinforcement learning technology applied to three-dimensional Unmanned Aerial Vehicle (UAV) air game maneuver decision-making often results in low utilization efficiency of training data and algorithm convergence difficulties. To address these issues, this study proposes an expert experience storage mechanism that improves the algorithm’s performance with less experience replay time. Based on this mechanism, a maneuver decision algorithm using the Dueling Double Deep Q Network is introduced. Simulation experiments demonstrate that the proposed mechanism significantly enhances the algorithm’s performance by reducing the experience by 81.3% compared to the prioritized experience replay mechanism, enabling the UAV agent to achieve a higher maximum average reward value. The experimental results suggest that the proposed expert experience storage mechanism improves the algorithm’s performance with less experience replay time. Additionally, the proposed maneuver decision algorithm identifies the optimal policy for attacking target UAVs using different fixed strategies.

Anomaly detection in multivariate time-series data remains challenging due to the complexity of temporal dependencies, non-stationary dynamics, and heterogeneous variable interactions. Existing methods often struggle to simultaneously capture correlations within and across different time-series components, limiting their capacity to uncover fine-grained causal relationships. This limitation is particularly acute in highly dynamic links, where rapid network fluctuations introduce substantial variability and uncertainty. To address these issues, we propose a causal discovery based timeseries anomaly detection method (CD-TAdetector). This method integrates a Granger causalitybased encoder-decoder structure with an attention mechanism to model causal dependencies across variables. To enhance robustness under dynamic conditions, we incorporate sparsification and temporal smoothing strategies. An attention-based module is further developed for interpretable root cause identification. Extensive experiments on three real-world datasets demonstrate that CD-TAdetector consistently outperforms state-of-the-art baselines in both anomaly detection and root cause identification, while providing improved interpretability for highly dynamic links.